3483
J . Am. Chem. SOC.1986, 108, 3483-3487
than the reactivity shown by the free flavins FAD and FMN themselves. Other flavoproteins show low reactivity with sulfite. It was therefore interesting that, unlike lipoamide dehydrogenase,= flavopapains lC, E , and X form reversible adducts with sulfite. The Kd constants for the formation of sulfite adducts of lA, lC, ZA,and 2C are 4, 15, 5 , and 16 mM, respectively (pH 7.5, 0.02 M Tris, 0.08 M KBr, 0.1 mM EDTA). The enhanced stability of sulfite adducts of the flavin-dependent oxidase enzymes has been correlated with protein stabilization of a flavin radical. Given the similarity in magnitude of the Kd values, it appears that the papain environment has little influence on the addition of sulfite to the N-5 position.
novel structure has been constructed that mediates the efficient oxidation of dithiols to disulfides. The better understanding of the diverse catalytic efficiency exhibited here and in previous studies remains a worthwhile goal. It appears that the different substitution pattern in lC, 2C, and 3C has created active sites of different shape, and for two classes of substrates, the active site of 1C appears to allow the closest approach of substrate to flavin in a manner that is along a productive reaction pathway. The versatility of the semisynthetic flavopapain catalysts is shown by their ability to oxidize dithiols and dihydronicotinamides. Future work will continue to focus on establishing general rules for optimizing the design of semisynthetic enzymes.
Conclusions The results of this work provide additional evidence that chemical modification of existing proteins is certainly a viable method for producing new catalysts. In particular, a catalyst of
Acknowledgment. The AVIV 60DS spectrometer was purchased with funds from National Science Foundation Grant PCM 84-00268. This work was supported by NSF Grant CHE-8218637 (E.T.K.) and by N I H Postdoctoral Fellowship F32 AM07523-01 (K.D.S.).
(23) Massey, V.;Muller, F.; Feldberg, R.; Schuman, M.; Sullivan, P. A.; Howell, L. G.; Mayhew, S. G.; Matthews, R. G.; Foust, G. P. J. Biol. Chem. 1969, 244, 3999-4006.
Registry NO. lA, 79127-38-1; lB, 79127-42-7; 2A, 63147-98-8; 2B,
68973-54-6; 3B, 101916-86-3; d,r-HS(CH,),CH(SH)(CH,),CONH,, 4265-09-2; d,l-HS(CH2)2CH(SH)(CH2)4CO2H, 751 6-48-5.
Praseodymium( 111) Transport across Phospholipid Vesicles in the Joint Presence of an Ionophore and a Fatty Acid+ Jean Grandjean and Pierre Laszlo* Contribution from the Institut de Chimie Organique et de Biochimie, UniversitZ de LiZge au Sart- Tilman, 4000 LiZge, Belgium. Received October 16, 1985
Abstract: Accelerated transport of Pr3+ions across phosphatidylcholinevesicles, as measured by 3’P NMR, occurs in the joint presence of an ionophore (X-537A) and a fatty acid, itself totally incapable of Pr3+transport. Several mechanisms are considered in turn, and they are ruled out by appropriate experiments. The explanation for the observed synergism is coupling of Pr3+ influx with H+(and Na+) efflux, in an “antiport” mechanism. The fatty acid present helps in this efflux: either because of its pK, difference with X-537A and/or because it serves as an uncoupler.
A remarkable synergism occurs in the transport of Pr3+across phosphatidylcholine vesicles: adjunction to lasalocid A (X-537A), a bona fide ionophore (A), of another ionophore (B), such as etheromycin or monensin, boosts the rate of transport.’ These initial observations were later generalized to include quite a number of paired ionophores of the polyether antibiotic type., Moreover, a similar synergism occurs when, instead of a second ionophore, a synthetic crown ether is incorporated in the lipid bilayer together with X-537A.’ All these findings applied to accelerated transport by a 2:l hybrid complex, in which the Pr3+ cation is coordinated to both ionophores A and B. All these findings applied to the influx of Pr3+ from the outside of the lipid bilayer to the inside of the vesicles. A common structural feature of all the species B that displayed, jointly with an ionophore A, synergistic transport of Pr3+ was the presence of a carboxylic group. Therefore, our working hypothesis was that, because of the necessary coupling between Pr3+ influx and H+ efflux for maintenance of electroneutrality, the presence of a ApK, between both acidic groups in the hybrid (Pr3+, A+B) complex could facilitate back-transport of protons and thus render Pr3+transport faster than in the “pure” (Pr3+,A,) or (Pr3+, B,) complexes.2 We test here this working hypothesis by studying Pr3+ transport in the joint presence of an ionophore (X-537A) and a fatty acid, which in itself is totally Presented in preliminary form at the 8th International Biophysics Congress, Bristol, England, July 29-August 4, 1984.
0002-7863/86/ 1508-3483$01.50/0
incapable of P$+ transport. We find indeed that the cation inward rate is markedly enhanced by the presence of the fatty acid. Materials and Methods Suspensions of vesicles were prepared as previously described2from a modification of a published proced~re.~The phospholipidic content was typically 30 mM, and the pH of the aqueous solution was adjusted with NaOH. Methanolic solutions of lasalocid A (Aldrich) and of the various fatty acids (Sigma or Aldrich) were added to the vesicle suspensions and incubated for at least 15 min at the temperatures of the various experiments. Palmitic acid ‘3C-labeledon the carboxyl carbon was incorporated for some of the experiments into the bilayer by sonication. The acidity constants K, have been determined by potenti~metry,~ with an ion analyzer,Orion Model 701A, in methanol-water mixtures (80:20, v/v), a medium that simulates very well the interface between a membrane and water? Ion-exchange chromatography (Dowex 50W ion ~
~
~~
(1) Grandjean, J.; Laszlo,P. Biochem. Biophys. Res. Commun. 1982,104, 1293-1297. (2) Grandjean, J.; Laszlo, P. J . Am. Chem. SOC.1984, 106, 1472-1476. (3) Bartsch, R. A.; Grandjean, J.; Laszlo, P. Biochem. Biophys. Res. Commun. 1983,117, 340-343. (4) Barenholz, Y . ;Gibbes, D.; Litman, B. J.; Goll, J.; Thompson, T. E.; Carlson, F. D. Biochemistry 1977, 16, 2806-28 10. ( 5 ) Rorabacher, D. B.; McKellar, W. J.; Shu, F. R.;Bonavits, S. M. Anal. Chem. 1971,43, 561-573. ( 6 ) Kauffman, R. F.; Taylor, R.; Pfeiffer, D. R. Biochemistry 1982, 21, 2426-2435.
0 1986 American Chemical Society
Grandjean and Laszlo
3484 J. Am. Chem. SOC.,VoI. 108, No. 12, 1986
Table I. Proton Longitudinal Relaxation Times T I (s) of Phosphatidylcholine Vesicles (30 mM) with and without Lauric Acid
concn lauric acid. mM
Figure 1. Example of deconvolution of an experimental line shape (-): subtraction of the high-field Lorentzian component (- due to vesicles a-)
devoid of internal Pr3+.
7
i
1.0
I
I
4
,
,
8
,
,
12
,
,
Nc
Figure 2. Influence of the chain length of fatty acids, as measured by the number of carbon atoms N,,on carrier-mediated(X-537A) transport of Pr3+as measured by the parameter S (Hz mi&). The determining
factor is incorporation of the fatty acid in the lipidic bilayer (see text). exchanger) provided lasalocid A under its acidic form. The transport rates are determined as in our earlier papers's2 by the NMR technique NMR spectrum of the suspioneered by the Bystrov group.' The pension of the vesicles, in the absence of the paramagnetic prascodymium(II1) ions, shows but one peak. The presence of in the external compartment shifts downfield the jlP resonance of the outer polar head groups. Penetration of the paramagnetic cations inside the vesicles induces also the downfield shift of the inner phosphate signal, whose shift rate thus monitors Pr3+transport. The ,IP NMR spectra have been obtained either with a Bruker WP-80 or with a Bruker AM-300 WB spectrometer, with nominal frequencies of 32.39 and 121.44 MHz. In the latter case, the dispersion of the chemical shifts makes it possible to distinguish, during the early part of each kinetic run, the vesicles without any paramagnetic P$+ ion inside from those having at least one such ion. Therefore, we could then determine the IlP chemical shift of the "inner" signal after subtraction of a Lorentzian line corresponding to those vesicles devoid of Pr'+. A typical example of this deconvolution procedure is shown in Figure 1. The longitudinal relaxation times T, have been measured with the AM-300 WB apparatus by using the (T-180°--t-900)n pulse sequence. Thirteen t values have been used to calculate the Tl's by fitting the experimental points to an exponential curve. The "C NMR spectra have been determined on the same high-field spectrometer operating at 75.43 MHz.
Results and Discussion Analysis of the kinetic data shows confirmity to a diffusive carrier mechanism.8 According to this valuable kinetic model, the progressive broadening of the inner resonance is consistent with an intermediate rate of mediator exchange between vesicles, relative to the rate of translocation. As we expected, the transport rate is increased significantly when fatty acids are present. We have investigated the effect of varying the length of the alkyl chain R of the saturated fatty acids RCOOH (Figure 2). A sigmoidal change is observed,plateauing to a maximum when the chain has 10 carbon atoms or more. The simplest interpretation involves the relative solubilities of these fatty acids in the aqueous and in the lipid phases: while butyric, n-caproic, and caprylic acids are soluble in water [miscible; 1.08 g/100 g and 0.068 g/100 g, respectively (Merck Index)], starting with n-capric acid, which is practically insoluble in water, the higher homologues (7) Bergelson, L. D.; Barsukov, L. I.; Dubrovina, N. I.; Bystrov, V. F.Dokl. Akad. Nauk SSSR 1970. 194.708-709. (8) Ting, D. Z.; Hagan, P.S., Chan, S.I.; Doll, J. S.;Springer, C. S.,Jr. Eiophys. J . 1981, 34, 189-216.
assgnt CH=CH, CH-OCO CHZOP, CHZOCO CH2N+ (CH3)3N+ CHzCO CHZC= (CH2)n CH,
0 0.221 0.442 5.35 0.40 (h0.02) 0.44 (f0.02) 0.47 (f-O.001)
6
4.34 0.35 (fO.01) 0.35 (fO.01) 0.37 (fO.01) 3.74 3.29 2.41 2.08 1.31 0.93
0.31 0.36 0.35 0.40 0.44 0.63
(fO.O1) (fO.01) (fO.O1) (fO.O1) (fO.01) (fO.01)
0.31 0.35 0.33 0.41 0.44 0.64
(fO.O1) (fO.01) (fO.01) (fO.O1) (iO.01) (fO.02)
0.31 0.39 0.35 0.43 0.45 0.69
(+0.01) (fO.01) (fO.O1) (iO.01) (f0.01) (fO.O1)
are lipophilic and will dissolve more readily in the lipidic bilayer. Similarly, this can be also expressed from the Hansch partition coefficients? Their logarithms are -3.2, -2.2, -1.2, -0.2, and +0.8 for the C4,CL,C8, CIo,and C12acids, respectively, paralleling the behavior observed here (Figure 2). To determine why incorporation of a fatty acid into a phosphatidylcholine vesicle increases the rate of influx of Pr3+ ions transported by X-537A from the oustide to the inside of the vesicle, we first checked that the observed effect is not due to major disruption of the membrane structure. A first possibility would be for insertion of fatty acid molecules into the lipidic membrane to increase the density of negatively charged groups at the surface of the vesicle: thus, more cations would be attracted, condensed, or bound into the electrical double layer, and the transport rate would be increased accordingly. But we could rule out such an explanation, at least as the predominant factor; in the Pr3+ concentration range investigated, the downfield shift of the outer phosphate signal (see Materials and Methods) is proportional to the mole fraction of bound Pr3+.Addition of lauric acid (lauric acid/phospholipid = 0.01)induces an extra shift of 5%. This would translate into a 5% increase in the transport rate, which is negligible by comparison to the observed increase of 250%. A second possibility is for insertion of fatty acids into the lipidic membrane to disrupt its structure to such an extent that Pr3+ transport will be boosted. We have investigated this factor by determining the 'H longitudinal relaxation times of the lipid bilayer as a function of the fatty acid concentration (Table I). Proton relaxation in membranes is dominated by dipolar interactions that are partially averaged by fast internal motions of the membrane lipids. Appropriate models describing the dynamics of the phospholipid chains have been reported in the The slow isotropic tumbling of the vesicles and the lateral diffusion of the lipids are also responsible for the partial averaging of the dipolar interaction^.'^.'^ A partial leveling of the proton relaxation times comes from a slow spin-diffusion process within the bilayer.15 The insertion of fatty acids in the lipidic phase induces perturbations in the mobility of the membrane,I6 and it should be reflected by proton relaxation times. The signal assingments have been taken from the 1iterat~re.l~ No significant change in the mobility due to the presence of the fatty acid is displayed by the results. The absence of a significant effect is due to the weak content of free fatty acid, ca. 1% instead of the 20% in the study by Bloom et a1.16 A third possibility is a chemical change in the composition of the membrane lipids induced by the presence of free fatty acid: (9) Hansch, C.; Lien, E. J. J. Med. Chem. 1971, 14, 653-670. (10) Lipari, G.; Szabo, a. J. Am. Chem. SOC.1982, 104, 4546-4549. (1 1) Pace, R. J.; Chan, S.I. J. Chem. Phys. 1982, 76,4217-4247. (12) Brown, M. F.J . Chem. Phys. 1984,80, 2808-2836. (13) Bloom, M.; Bumell, E. E.; MacKay, A. L.; Nichol, C . P.;Valic, M. I.; Weeks,G. Biochemistry 1978,17, 5750-5762. (14) Bocian, D. F.;Chan, S.I. Annu. Rev. P h p . Chem. 1978,29,307-335. (15) Ellena, J. F.;Hutton, W. C.; Cafiso, D. S.J . Am. Chem. SOC.1985, 107, 1530-1537. (16) Pauls, D. P.;MacKay, A. L.; Bloom, M. Biochemistry 1983, 22, 6101-6109. (17) Kurcda, Y.; Kitamura, K. J . Am. Chem. SOC.1984, 104, 1-6.
J . Am. Chem. Soc., Vol. 108, No. 12, 1986 3485
Pr" Transport across Phospholipid Vesicles Table 11. pH Dependence of the Pr3+ Transport Ratea DH slow ( i o ) p concn lauric acid, mM ~
6.21 6.40
6.51 6.62
1 .oo (f0.02) 1.61 (10.08) 1.31 (f0.03) 1.56 (f0.05) 2.20 (f0.08) 3.66 (f0.28) 7.20 (f0.40) 8.60 (f0.37) 1.59 (f0.09) 8.06 (f0.50) 1.70 (10.071 6.51 i a 2 5 i
0.992 0.991 0.997 0.991 0.995 0.987 0.994 0.995 0.988 0.993 0.994 0.996
0 0.29 0 0.07 0.15 0.29 0.44 0.58 0 0.29 0 0.15
'Key: [Pr3+] = 1.03 mM; [X-537A] = 0.11 mM: P = linear correlation coefficient (8-10 points); T = 305 K.
addition of RCOOH molecules might cause hydrolysis of the ester bonds in the phosphatidylcholines with release of complex higher fatty acids, some of which might act as Pr3+ carriers. But this would cause a comparable increase in the cation-transport rate with all the acids studied, contrary to what is observed (Figure 2). In addition we could disprove this possibility experimentally. We have examined for this purpose 13C NMR spectra for 13COOH-enrichedpalmitic acid incorporated into the vesicular membrane. Acyl transfer between palmitic acid and the phospholipids present would increase the intensity of the ester resonance (6 174.0) and conversely deplete the free carboxylic groups (6 175.5). Even after 1 week, the intensity ratio of these two signals in fact remained unchanged. In order to ascertain the mechanism for the observed synergism in the presence of fatty acids (N, > 10, Figure 2), we elected lauric acid (Clz) as a representative case for further study. First, we checked that it does not belong in the class of lipidic compounds such as cardiolipin for which ionophoretic properties have been claimed:I8J9 under the conditions of our experiments, lauric acid does not transport Pr3+. Yet, in association with lasalocid A, lauric acid gives rise to an enhanced transport rate (Figure 2). This is a property of the free acid. By contrast to lauric acid, the methyl laurate ester does not increase translocation of the paramagnetic cation mediated by lasalocid A: presence of the carboxylic or the carboxylate group is necessary to enhance the influx rate of Pr3+;the same is true with higher fatty acids, which are known to be extensively incorporated into the bilayer.'6u20 Furthermore, the pH dependence of the observed enhancement implies, as expected, intervention of the carboxylate group (Table 11). Added insight into the mechanism of this synergism is obtained from the dependence of the cation inward rate upon the concentrations of lasalocid A and lauric acid. At pH 6.5, the rate is second order in lasalocid A [correlation coefficient 0.991 (five points) (Table III)] exactly as had been reported for the ionophore alone,21Jand the transport rate is first order with respect to lauric acid: correlation coefficients are 0.991 (five points) at pH 6.0 (Table 111) and 0.988 (four points) at pH 6.4 (Table 11). If penetration of the complex into the membrane is rate determining:, the role of lauric acid could be that of a shield around the Pr3+-(lasalocid A), complex to provide it with better hydrophobicity. Indeed, in liquid-liquid extraction by carboxylic crown ethers, 2: 1 stoichiometries (crown ether/univalent or divalent cation) have been proposed for the cationic complex involved in the metal e ~ t r a c t i o n :the ~ ~ second complexant molecule serves (18) Cullis, P. R.; De Kruijff, B.; Hope, M. J.; Nayar, R.; Schmid, S. L. Con. J . Biochem. 1980, 58, 1091-1100.
(19) Gerritsen, W. J.; De Kruijff, B.; Verkleij, A. J.; De Gier, J.; Van Deenen, L. L. M. Biochim. Biophys. Acta 1980, 598, 554-560. (20) Parman, Y. I.; Wassal, S.R.; Cushley, R. J. J. Am. Chem. Soc. 1984, 106, 2434-2435. (21) Fernandez, M. S.; Celis, H.;Montal, M. Biochim. Biophys. Acra 1973, 323, 6OC-605. (22) Degani, H.; Simon, s.; McLaughlin, A. C. Biochim. Biophys. Acra 1981,646, 320-328. (23) Fyles, T. M.; Withfield, D. M. Can. J . Chem. 1984, 62, 507-517.
Table 111. Pr3+ Transport Rate vs. Lauric Acid or Lasalocid A Concentration slow (fu1 0.58 0.73 0.74 1.01 1.85
(f0.03) (f0.03) (f0.05) (f0.07) (f0.09)
0.65 ( f 0 . 0 5 ) 2.47 (fO.O1) 2.60 (f0.09) 6.30 (f0.23) 9.28 (f0.551
Pa
concn. mM
Lauric Acidb 0.992 0.996 0.994 0.992 0.996
0.13 0.20 0.20 0.27 0.40
X-537AC 0.990 0.996 0.994 0.997 0.993
0.04 0.08 0.08 0.12 0.16
O p = linear correlation coefficient (8-10 points). bpH 6.0; [Pr"] = 1.12 mM; [X-537A] = 0.12mM; T = 305 K. 'pH 6.5; [Pr'+] = 1.03 mM; [lauric acid] = 0.27 mM; T = 305 K.
as a counterion that can be displaced by picrates for instance.
Nevertheless, for the kinetic results analyzed here, it seems unrealistic to assume involvement of a quaternary complex in the slow step. Conversely, if the rate-determining step is diffusion of the Pr3+-transporting complex through the bilayer, a sizable increase in the lipophilicity of this complex might be expected to alter the activation energy of the overall process. We have determined Arrhenius activation energies for the translocation of Pr3+ in the presence and in the absence of lauric acid: 92 f 8 and 88 f 8 kJ mol-', respectively. These activation energies, similar in magnitude to those obtained with other are equal within experimental error! Cation transport could also be governed by steps other than diffusion of the ~ o m p l e x . ~ Thus, ~ - ~ ~it should be recalled that Pr3+influx destroys intravesicular electroneutrality. The chloride anion can indeed permeate the vesicles, but at a pH of 6.2 (ApH 0), its permeability coefficient is ca. 5 X lo-', ~ m / s , ~Le., I too low. Furthermore, it seems to be due to the transport of HCl.32 The passive permeability of Na+ is 1.2 X cm/s." True, the sodium efflux can be mediated by the ionophore itself but much less efficiently than for the paramagnetic ion. Therefore, neither C1- nor Na+ fluxes are sufficient to restore internal neutrality under the time scale of our experiments. By contrast a proton antiport mechanism appears to be operative. In the absence of lauric acid, the ionophore alone provides a pathway for proton countertransport. The proton is translocated as HA2- species where A- is the anionic form of the ionophore. It results from the association of the undissociated ionophore HA and its form A-?33 Thus,in order to maintain internal neutrality, more lasalocid A molecules are required for the outward flux of protons than for the influx of Pr3+. The question arises of whether the added lauric acid could provide an extra pathway for proton countertransport. A first line of evidence comes from the increase in the transport rate at an higher pH (Table 11). It indicates that the anionic conjugate base is the active ingredient. Such a possibility has been indirectly (24) Degani, H.; Lenkinski, R. E. Biochemistry 1980, 19, 3430-3435. (25) Hunt, G . R. A.; Tipping, L. R. H.;Belmont, M. R. Biophys. Chem. 1978,8,431-435. (26) Grandjean, J.; Laszlo, P. In Physico-Chemistry of Transmembrane Morion; Spach, G., Ed.;Elsevier: Amsterdam, 1983; pp 289-295. (27) Degani, H.Biochim. Biophys. Acra 1978,509, 364-369. Chiu, V. C. K.; Watson, B. Arch. Biochem. Biophys. (28) Haynes, D. H.; 1980, 203, 73-89. (29) Lauger, P.; Benz, R.; Stak, G.; Bamberg, E.; Jordan, P. C.; Fahr, A,; Brock, W. Q.Reu. Biophys. 1981, 14, 513-598. (30) Laprade, R.; Grenier, F.; Lapointe, J.-Y.; Asselin, S. J. Membr. Biol. 1982,68, 191-206. (31) Hauser, H.;Oldani, D.; Philipps, M. C. Biochemisfry 1973, 12, 4507-45 16. (32)-Kornberg, R. D.; McMamee, M. G.; McConnell, H.M. Proc. Nafl. Acad. Sci. W.S.A. 1972, 69, 1508-1513. (33) McLaughlin, S.; Eisenberg, M. Annu. Reu. Biophys. Biwng. 1975, 4, 335-366. (341 Celis, H.; Estrada, 0. S.; Montal, M. J. Membr. Biol. 1974, 18, 187-199.
Crandjean and Laszlo
3486 J. Am. Chem. Soc., Vol. 108, No. 12, 1986 Scheme I
B
A
X -537A-COOH Iaur- COO-,
/ 1
0
1
"talc
‘“\
I n n e r Interface
/
Figure 3. Calculated proton transport rates vs. lauric acid and lasalocid A concentrations, assuming k2/kl = 0.05: (A) [X-537A]= 0; (B) [X-537A]= 0.11 mM; (C) [X-537A]= 0.22 mM.
concentration used in the transport experiments (Table 11). With such an assumption, the rate Vdd has been plotted vs. the fatty acid and ionophore concentrations (Figure 3). With an ionophore proposed already for myristic and oleic acids.35 Picric acid, a concentration of 0.1 1 mM, the calculated curve does not deviate well-known p r o t ~ n o p h o r e ? ~also * ~ increases ~ Pr3+influx mediated significantly from a straight line except at very low lauric acid by lasalocid A,2 which shows that proton efflux is a limiting step concentrations. This simulation is in full agreement with our for cation influx. experimental results. As expected, due to the quadratic effect We asked how we could explain the first-order dependence in of the lasalocid c o n c e n t r a t i ~ n ,at ~ ~higher . ~ ~ ionophore levels the lauric acid, observed for Pr3+ transport, when mediated by the curvature occurs at the highest fatty acid concentrations. same amount of X-537A. Proton transport can occur by three The above assumption of a rate law first order in carboxylic different processes: passive transport,38transport by the lasalocid acid is consistent with the data in Table 11: the kinetic order in ionophore or transport by lauric acid and its conjugate base. H+, based on these few points, appears as -0.6 with lasalocid alone In small unilamellar vesicles, the net permeability to protons is and between -2 and -3 with added lauric acid. As indicated by 5 X lo-’ C ~ / S .The ~ ~uncoupler properties of the i ~ n o p h o r e ~ ~ f ~ the reviewer, these are very reasonable results, as the electroneutral render negligible in its presence the passive process. We have transport by (lasalocid-H-lasalocid-) dimers should have inverse estimated a lower limit for the proton efflux due to the presence first order in protons,” while a lauric acid cycle that achieves of a fatty acid as follows: the 13Cresonance of 13COOH-enriched electroneutrality should show inverse cube order in protons. palmitic acid incorporated into the bilayer is pH dependent. If We had earlier postulated that the ApKa intrinsic to the hybrid palmitic acid molecules were to crisscross the bilayer at a rate complex formed between two ionophores might help in “pulling” slow on the NMR time scale, and in the presence of an adequate protons through the water-membrane interface? Obviously, such pH gradient across the membrane, one should see two resonances a process remains possible here. We can visualize such a kinetic due to the inner and the outer carboxyl groups. At pH 6.2 (ApH effect in the (admittedly improbable, but this would provide the 0), the carboxyl signal occurs at 176.2 ppm. When the external lowest activity energy pathway) situation having the least basic pH is increased to 8.35 by addition of sodium hydroxide, one single carboxylate group sitting at the interface and in rather close line is observed at 178.7 ppm: on the basis of the chemical shift proximity to the most basic carboxylate group, the latter buried difference, the flip-flop rate of palmitic acid between the two more deeply in the lipophilic membrane environment (Scheme compartments is at least 2.5 X lW5s-l. Taking into account the I). As soon as a proton would attach itself to the least basic dimensions of small phosphatidylcholinevesicles?’ the resulting COO-, its further jumps acrass the more basic COO-, its further permeability c o e f f i ~ i e n tis~ .2.5 ~~X cm/s, much higher than jumps across the more basic COO- would become highly probable. that for the passive transport of protons (see above). (Note that Hence, such a mechanism may well facilitate penetration of absence of chloride ions and an internal compartment more acidic protons from the inner compartment into the lipid bilayer, and than the external are required to avoid passive permeation of this would accelerate transport, as effected and seen by Pr3+. protons.40) No value for proton translocation in the presence One last question that should be addressed is whether a pK, of lasalocid A has been reported. However, the proton transport difference does indeed exist in the lauric acid-lasalocid A system. rate constant mediated by A23 187, another powerful ionophore, We have determined the two acidity constants for these carboxylic is 28 s-’.~~ acids in methanol-water mixtures (80:20, v/v) chosen to mimic The proton transport rate coupled to the Pr3+ influx can be the membrane interface.6 The corresponding pK,’s are 6.98 f expressed as a sum of two contributions 0.02 and 5.22 f 0.02 for lauric acid and for X-537A, respectively. The value for the antibiotic agrees well with a pK, = 5.0 f 0.2 Vcslcd = k,[X-537Al2 + k2[lauric acid] estimated from the cation transport mediated by this ionophore where kl and k2 are the apparent rate constants for proton fluxes at different pH values.27 Furthermore, the pKa of lauric acid in the presence of lasalocid A and of lauric acid, respectively. agrees very well with the values reported for other fatty acids Taking into account the above estimates for the rate constants incorporated into phospholipid vesicles.42 It appears to further together with the used concentrationsin our experiments, the last justify the choice of the above solvent mixture as a model for the term is obviously predominant. The second term is about 3 orders membrane interface. of magnitude greater than the first at the highest lauric acid Using fluorescent intravesicular pH probes such as sodium 1,3,6,8-pyrenetetrasulfonatehas given us a direct proof of the (35) Cunaro, J.; Weriner, M. W.Biochim. Biophys. Acta 1975, 387, coupling between Pr3+ transport and H+ countertransport. 234-240. Following entry of the paramagnetic cations, the intravesicular (36)Deber, C. M.; Young, M. E. M.; Kun, J. T. Biochemistry 1980,19, pH increases gradually. Joint addition of the ionophore and of 6194-6198. lauric acid increases significantly this proton efflux, in full (37)McLaughlin, S.;Dilger, J. P. Physiol. Rev. 1980, 60, 825-863. (38) Cafiso, D. S.:Hubbell, W. L. Biophys. J . 1983.44. 49-57. agreement with the mechanism earlier postulated.2 These mea(39)Mimms, L.T.; Zampighi, G.; Nozaki, Y.; Tanford, C.; Reynolds, J. A. Biochemistry 1981, 20, 833-840. (40)Sanson, A.,personal communication. (41) Kolber, M. A.; Haynes, D. H. Biophys. J. 1981, 36, 369-391.
(42)F’tak, M.; Egret-Charlier, M.; Sanson, A.; Boulousa, 0. Biochim. Biophys. Acta 1980, 600,387-397.
J . Am. Chem. SOC.1986, 108, 3487-3498
3487
surements will be reported in detail elsewhere. Finally, one could question the physiological relevance of these experiments and observations. However, Pr3+is an often-used Ca2+substitute!' Also,and although free-acid acids are minor components of biological membranes, exogeneous additions alter several membrane-mediated cellular functions such as cell p e r m e a b i l i t ~or ~ ~activity of membrane-bound
Furthermore, free fatty acids can be formed within the membrane by the action of phospholipases. To sum up, the increase of Pr3+ translocation across phosphatidylcholine vesicles by lasalocid A, as induced by the single adjunction of lauric acid, finds its most likely explanation in participation of the fatty acid in the proton countertransport. Such a cation-proton antiport process could be used to model cationproton exchanger systems in biological m e m b r a n e ~ . ~ ~ - ~ O
(43) Mikkelson, R. B. In Biological Membranes; Chapman, D., Wallach, D. F. H., Eds.; Academic: London, 1976; pp 153-190. (44) Schram, N.; Eisenkraft, B.; Barkai, E. Biochim. Biophys. Acta 1967, 135, 44-52. (45) Schmalzig, G.; Kutchera, P. J . Membr. Biol. 1982, 69, 65-76. (46) Rhoads, D. E.; Peterson, N. A.; Raghupathy, E. Biochemistry 1982, 21,4782-4787. (47) Pieper, G. M.; Salhany, J. M.; Murray, W. J.; Wu, S.T.; Eliot, R. S.Biochim. Biophys. Acta 1984,803, 229-240.
Registry No. X-537A, 25999-31-9; Pr, 7440-10-0; H ' , Na, 7440-23-5; lauric acid, 143-07-7.
12408-02-5;
(48) Bassalina, M.; Damanio, E.; Leblanc, G. Biochemistry 1984, 23, 5288-5294. (49) Nakamura, T.; Tokuda, H.; Unemoto, T. Biochim. Biophys. Acta 1984, 776, 33C-336. (50) Kinsella, J. L.; Aronson, P. S.Am. J . Physiol. 1980, 238, 461-469.
ENDOR of the Resting State of Nitrogenase Molybdenum-Iron Proteins from Azotobacter vinelandii, Klebsiella pneumoniae, and Clostridium pasteurianum: H, 57Fe,95Mo, and 33SStudies Ronald A. Venters,'. Mark J. Nelson,lb Paul A. McLean,lCAnne E. True,ld Mark A. LRvy,le Brian M. Hoffman,*ld and William H. Orme-Johllson*lc Contribution from the Department of Chemistry, Northwestern University, Evanston, Illinois 60201, and Department of Chemistry, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139. Received August 19, I985
Abstract: Electron nuclear double resonance (ENDOR) studies of native and isotopically enriched MoFe proteins hold the promise of individually characterizing every atom of the catalytically active FeMo-co cluster of the nitrogenase MoFe protein. and 33SENDOR measurements in a comparison of the MoFe protein isolated from This report presents 'H,57Fe,95Q7M~, the three titled organisms, Avl, Kpl, and Cpl. We have examined in detail single-crystal-like 57Feresonances from at least five distinct iron sites in each of the three enzymes, revising somewhat our earlier assignments. The analysis incidentally gives the electron spin zero-field splitting parameters to high precision. 95MoENDOR measurements for Cpl and Kpl give 95Mo hyperfine and quadrupole coupling constants. They indicate that a single molybdenum is integrated into the MoFe spin system and that the molybdenum is most plausibly viewed as being in an even-electron state, which may be assigned provisionally as unsymmetrically coordinated Mow. The observation of an exchangeable proton or protons from each protein source suggests a site on the cluster accessible to solvent and perhaps containing H20or OH-. Cpl enriched in ')S gives the first observed ENDOR signals from this nucleus. The resonances from 33Sare assignable to the inorganic sulfur because, it is argued, the FeMo-co cluster must be bound to the protein primarily, if not exclusively, by residues other than cysteinyl.
Nitrogenase comprises two extremely oxygen-sensitive metalloproteins, the iron protein (Fe protein) and the molybdenumiron protein (MoFe protein).2 The Fe protein accepts electrons from oxidative processes operating at or below -0.4 V and then specifically passes them on to the MoFe protein for use in the reduction of dinitrogen and other substrates.' The MoFe protein (1) (a) Northwestern University. Present address: Benger Laboratory, E. I. du Pont de Nemours and Co., Waynesboro, VA 22980. (b) Massachusetts Institute of Technology. Present address: Experimental Station, E. I. du Point de Nemours and Co., Wilmington, DE 19898. (c) Massachusetts Institute of Technology. (d) Northwestern University. (e) Massachusetts Institute of Technology. Present address: Smith Kline and French Laboratories, Philadelphia, PA 19101. (2) The following abbreviations are used: Avl, Kpl, and Cpl, the molybdenum-iron proteins, respectively, from A . vinelandii, K.pneumoniae, and C. pasteurianum; MoFe, molybdenum-iron protein; FeMc-co, molybdenumiron cofactor; ENDOR, electron nuclear double resonance. (3) (a) Orme-Johnson, W. H. Annu. Rev. Biophys. Chem. 1985, 14, 419-459. (b) Nelson, M. J.; Lindahl, P. A.; Orme-Johnson, W. H. In Advances in Inorganic Biochemistry; Eichorn, G . L., Marzilli, L. G., Eds.; Elsevier: New York, 1981; 1-40. (c) Mortenson, L. E.; Thorneley, R. N . F. Annu. Rev. Biochem. 1979, 48, 387-418. (d) Brill, W. J. Microbiol. Rev. 1980, 44, 449-467. (e) Yates, M. G. Biochem. Plants 1980, 5, 1-64. (f) Robson, R. L.; Postgate, J. R. Annu. Rev. Microbiol. 1980, 34, 183-207.
0002-7863/86/ 1508-3487$01S O / O
as isolated from Azotobacter vinelandii (Avl)? Clostridium pasteurianum ( C P ~ ) ,and ~ . ~Klebsiella pneumoniae ( K P ~ is) ~an a2j32tetramer with molecular weights of approximately 240 000 for Avl and 220000 for both Cpl and Kpl. Chemical analysis of the MoFe protein has shown it to contain 2 mol of molybdenum, approximately 30 mol of iron, and approximately 30 mol of acid-labile sulfur/mol of peptide component.* Accurate chemical and spectroscopic assignment of these molybdenum, iron, and sulfur atoms into individual clusters and investigation of the structure of those clusters constitute much of the recent progress in the study of nitrogenase. In particular, (4) Shah, V. K.; Davis, L. C.; Brill, W. J. Biochim. Biophys. Acta 1972,
256. 498-5 ---. - - -1 -1. (5) Nakos, G.; Mortenson, L. E. Biochim. Biophys. Acta
1971, 229, 431-436. (6) Zumft, W. G.; Hase, T.; Matsubara, H. In Molybdenum Chemistry of Biological Significance; Newton, W. E., Otsuka, S.,Eds.;Plenum: New York, 1980; p 59-72. (7) Eady, R. R.; Smith, B. E.; Cook, K. A,; Postgate, J. R. Biochem. J . 1972, 128, 655-675. (8) Davis, L. C.; Shah, V. K.; Brill, W. J.; Orme-Johnson, W. H. Biochim. Biophys. Acta 1972, 256, 512-523.
0 1986 American Chemical Society
